This is technically not true. Yes, Pulleys & Belts are great for multiplying motor torque, however, there is a limit to how much torque can be transferred by a belt drive. The belt is a torque bottleneck & also introduces many other design constraints & maintenance requirements.
When it comes to performance what is most important is torque at the wheel. This is why car enthusiasts & engine tuners always measure torque at the wheel with the car/bike strapped to a dyno, the dyno results represent total power minus transmission losses.
For example, Take a 9mm wide, 5mm pitch belt drive on a normal satellite drive train on an electric skateboard, you can transfer a max of 2Nm of torque into the wheel at 1000rpm.
After that, you get belt slipping & reduced service life of your drive train (I dont have simulations for brake force, I believe peak torque for braking is actually much greater so you get more slipping). It also doesn’t really matter if you have an idler either, eventually you reach the torque transfer limits. In fact, the only way to increase the torque transfer is to increase the size of the motor pulley & your belt width. This reduces reduction & increases losses, eventually you end up nearly-having a direct drive system! However, you still have transmission losses due to belt friction/resistance.
Now, Install a hub motor & the power is already at the wheel, with the new R-SPEC GHOST Direct Drive simulations, we see (3X) 6nm of torque at the wheel at 1000RPM - the torque is also delivered at 55% HIGHER BOARD SPEED!
When you develop a motor for a specific application you will always get better results than when adapting a motor with a transmission system. Transmission systems are just not needed in the new era of motor technology, they are inefficient at transferring torque, they are more complex to assemble, prone to damage from debris, need regular maintenance & have a higher running cost due to the lifespan of the belts.
Direct drive systems are appearing everywhere now… there is a reason for this!